Development of single-electron reservoir compuiting

2023 Fiscal Year Final Research Report

• Project/Area Number: 19H02545
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 28020:Nanostructural physics-related
• Research Institution: Yokohama National University
• Principal Investigator: OYA Takahide 横浜国立大学, 大学院工学研究院, 准教授 (30432066)
• Project Period (FY): 2019-04-01 – 2024-03-31
• Keywords: 単電子回路 / リザーバコンピューティング

Outline of Final Research Achievements

The purpose of this research is to develop a circuit architecture for implementing the reservoir computing (RC) model in a single-electron device. RC's greatest feature is a method for setting connections and coupling weights between neurons contained in a part called the reservoir, which do not require settings and methods that have needed to be fully considered in general neural networks.
In this study, I aimed to construct a new single-electron information processing mechanism that can solve the problems of the number of wires, learning (changing the weight of synaptic connections), and noise utilization, which are issues in single-electron circuit research, and found the possibility of such a mechanism.

Free Research Field

ナノエレクトロニクス

Academic Significance and Societal Importance of the Research Achievements

本研究では特に、回路の大規模化で大きな問題となる配線数について、リザーバコンピューティングを実装することで課題解決と情報処理の両立が可能なことを見出したことが大きいと考える。
今回は単電子回路を対象として、その可能性について研究を遂行したが、他のナノ・量子デバイスや、いわゆる半導体集積回路においても同様の手法が取れる可能性があり、回路の大規模化に伴う配線数の増加に悩まされる情報処理デバイス全般に対して、ある程度の貢献ができると言える。